Xref: utzoo comp.dsp:1755 sci.electronics:20690
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!cs.utexas.edu!sun-barr!newstop!sun!amdcad!jetsun!pyramid!prls!max
From: max@prls.UUCP (Max Hauser)
Newsgroups: comp.dsp,sci.electronics
Subject: What's real answer re 180-degree phase-shift algorithm
Summary: For amusement purposes only
Message-ID: <51199@prls.UUCP>
Date: 6 Jun 91 00:27:43 GMT
Reply-To: max@prls.UUCP (Max Hauser)
Followup-To: comp.dsp
Organization: Philips Research Labs, Sunnyvale, California
Lines: 48

People keep asking me this.  One sensible correspondent writes:

    So what ... IS the answer???  this isn't a challenge, I'm
    really curious.  

My object was to illuminate the game, not to jump in and play it.  (Often
I've seen these hilarious threads and wished later that I had saved one
of them as an example, useful for years.  This time I finally did so.)

It is actually simple enough to exactly, verifiably, comprehensively answer
the question as asked, even in its vagueness, if you know the subject.  A
slightly simpler answer, capturing the *essence* of the situation if not
all of the nuance, is that you multiply by -1.  A few respondents said so
promptly, and were buried in the noise and/or attacked.  (I imagine that
some readers will disbelieve this answer; in classic Usenet protocol they
would then ask me to "prove" it to their satisfaction, in language of their
liking -- while standing on one foot, if you will.  On the Usenet if you
actually know what you are talking about you not only had to work harder to
get there, but you also are expected to bear all the burden of proof.)
 
    (I don't profess to know electronics beyond the hobby level ...
    kept up on the easy stuff, which is to say digital electronics, but even
    undergrad-level Fourier analysis is is WAY beyond me.) 

Nothing wrong with that.  Most technical subjects are way beyond me, too,
certainly.  Of course, neither of us (nor most people) is in the habit of
holding forth, with an air of authority, on an international forum, with
uncontrolled circulation possibly in the thousands, about subjects that are
way beyond us (let alone WAY beyond us :-).  And we certainly take some
responsibility for knowing what is and is not beyond us.
    
    One poster claimed that "phase shift IS defined even for
    nonperiodic signals", but I can't imagine how.  ???

I could undertake to painstakingly explain this and other principles, but
it's really separate from the point of my postings.  If you know signals
then you know about "phase" in nonperiodic signals and all of the
implications of "180 degrees" and why "any signal with a Fourier transform
means any periodic signal" is unintentionally humorous.  If you don't then
usually it would simply not occur to you to hold forth on an international
forum etc. etc.  In this regard the Usenet accomplishes "noise enhancement."


PS My posting on this subject engendered a number of mails from long-lost
co-workers, classmates, and prospective correspondents, all asking "Are you
the Max Hauser who ..."   I don't know why this should happen suddenly (I
didn't just start posting), but in case anyone else is planning to ask that
question I should advise that so far the answer has always been yes.